The Kinesis platform provides a suite of analytical and modelling tools for working with urban data. The platform provides a common language for describing cities and the environment, to drive built-in or user-supplied models. With access to these tools, domain experts are able to interact with advanced models in a language they understand, collaborating on building scenarios and analysing results.

This article explains the central concepts in the platform and how they relate to each other:

In Kinesis, analysis and collaboration happen in workspaces. Each workspace has a particular focus that might represent a development project, plan, change in policies, etc.

A workspace contains scenarios, apps, boards, and other datasets. Users interact with apps in a workspace by defining locations and their attributes in scenarios as inputs, and visualising their outputs in boards.

When a workspace is first created, it starts off with an initial Kinesis platform app that provides an overview of the scenarios, and with a Baseline scenario that they then define. Depending on the purpose of the workspace, a user might create a number of scenarios to compare.

Apps get added to a workspace. The workspace owner determines what apps can get added to that workspace. When an organisation has a licence to an app, they can add this app to any and all of their workspaces.

Kinesis allows users to define and understand urban scenarios. For example, users can model how a region develops over time or how different policy measures affect the built environment. At its most basic, a scenario represents a specific possibility. A scenario consists of a set of locations with some attributes, which form inputs to your apps.

Scenarios exist within a workspace and are independent of each other. Scenarios can have entirely different locations, or even the same locations with different attribute values.

A location is any precisely specified point or area on Earth's surface. In a scenario, locations define the area or place of interest for the analysis. Each location belongs to a layer that represents a particular level of geography such as buildings, lots, suburbs, cities, states, and so on. Layers are used to define the relationship between locations. Details about these locations are added as attributes.

An app is a packaged collection of data and/or models aimed at solving a specific urban challenge. Apps provide expertise to the end user to help guide their decision making process. Apps are implemented as a set of models, normally written in Python, and a series of configuration that helps the platform understand what the app needs and how to best visualise the outputs.

When an app is added to a workspace, it also brings in any dependencies it might have like other apps or resources, as well as any boards that will visualise outputs.

The app takes the scenarios in that workspace and any other resources it needs as input.

Every time there is a change in the inputs, work is triggered to calculate outputs. These outputs are published as tables that are used by boards to visualise them.

Boards are a collection of blocks — visualisations, text, etc. that can be created by users or provided by apps as a starting point, to guide a user's thought process in solving a particular urban challenge.

Next up: How users interact with apps